Kile works in KDE window manager in Ubuntu! and the video cam, wireless, and microphone divices all work!

I got a surprise when I tried KIle in the kdm (KDE window manager or desktop). Kile latex IDE works fine in it and there are no more annoying messages about creating a default factory method. I can now insert math symbols selected from the left panel.
Its still the same 2.1.0 version of Kile. So I suspect that there were wrong settings for GNOME when I was still in Ubuntu 11.04.

This may now be my default window manager in my laptop while my ACER ONE still uses unity desktop. The system response is perceptibly fast!!

I went back to GNOME and I pleasantly discovered that I got my video and microphones and wireless working! Of course I updated my packages after upgrading to the Ocelot Ubuntu distribution. This may have fixed any issues with the softwares in the first upgrade.

Scite 2.03 is immovable on the desktop!

After trying out drpython, I installed scite and run it. But bad luck struck under Ubuntu. The editor screen cannot be moved! It was immobile on the left side of the screen and I cannot strangely zoom it. So another editor bit the dust in our Ubuntu natty narwhal OS powered desktop!

Note: Feb. 19, 2012. I looked at scite and since I have upgraded to Ubuntu oneiric ocelot OS version, Scite is working properly.

Dr Python is a nice lightweight Python editor but there is something wrong with screen rendering.

The following image shows that the screen rendering of drpython leaves much to be desired.

You have to scroll down and up so that the screen will be properly rendered.

Octobet 30, 2011: The funny and erroneous black bars have disappeared! after a few updates!

Alternative sites for information on free-software.

I will start collecting some urls which are good sources for free-software. The first esteemed entry is

1. Free Software Foundation Directory
2. Feeding the Snake
3. PyPi: Python directory of applications.Almost all are FREE!
   

As usual, most posts I have written are works in progress.

Installing Gretl in Ubuntu Natty

I get segmentation fault on startup of gretl. Here are some steps I took to make it run again
on my 64 bit ACER laptop. Download the 1.9.5 source code by clicking on

sourceforge: gretl

The actual download site link I got was

http://nchc.dl.sourceforge.net/project/gretl/gretl/1.9.5/gretl-1.9.5.tar.bz2

and if you prefer the command line(with a network connection!)

wget -ct0 http://nchc.dl.sourceforge.net/project/gretl/gretl/1.9.5/gretl-1.9.5.tar.bz2,

but that is not recommeded as the actual download site may be too busy!


Now go to the directory where the bzip file is, and unpack it with

tar xjvvf gretl-1.9.5.tar.bz2

then, goto the gretl-1.9.5 directory, and perform the standard

./configure; make; make install

incantation to compile and install from source files.


Here are some errors I got when compiling the sources:

../lib/src/gretl_fft.c:22:19: fatal error: fftw3.h: No such file or directory

Fix: install libfftw3 development files.(libfftw3-dev)

../lib/src/monte_carlo.c:40:17: fatal error: gmp.h: No such file or directory

Fix: install gmp development files. (libgmp3-dev)

Typing gretl on the command line does not do anything! So we type gretl again and the TAB key:
gretl
gretl gretlcli gretl_x11

Trying out gretlcli,

toto@toto-Aspire-4520:~/Downloads/gretl-1.9.5$ gretlcli
gretlcli: symbol lookup error: /usr/lib/libgretl-1.0.so.0: undefined symbol: omp_get_wtime

So back to installing omp (open MP). Fix: install libgomp-1-dbg

Darn, I still get the error!

gretlcli: symbol lookup error: /usr/lib/libgretl-1.0.so.0: undefined symbol: omp_get_wtime

So I thought that that we remove some remaining packages which are not in use anymore.

sudo apt-get autoremove.

Now we get a sweet response from grelcli:

gretlcli
gretl version 1.9.5
Copyright Ramu Ramanathan, Allin Cottrell and Riccardo "Jack" Lucchetti
This is free software with ABSOLUTELY NO WARRANTY
Current session: 2011-09-09 09:45
gretl is process 20218, '/usr/local/bin/gretlcli'
gretldir is maybe '/usr/local/share/gretl/'?

"help" gives a list of commands
Type "open filename" to open a data set
? ^C

and typing gretl now give us a more familiar qui (from our experience with previous versions).

I ahve upgraded to the newest Ubuntu distribution Oneiric Ocelot and gretl seems to work ("seemst to") because I have to spend more time with this econometrics software.

Links to Online Documentation

I am starting to gather links to manuals, tutorials  of free software.

1. Maxima Computer Algebra System.
   Manual: http://maxima.sourceforge.net/docs/manual/en/maxima.html
2. Python Programming Language.
   Main Documentation Page: www.python.org/doc/
   Reference: docs.python.org/reference/   
3. R statistical  software.
   Index to manuals: http://cran.r-project.org/manuals.html
4. Gretl Econometrics Software.
   Gretl-Guide: http://ricardo.ecn.wfu.edu/pub/gretl/manual/en/gretl-guide.pdf
5. Scipy and Numpy.
   http://docs.scipy.org/doc/

Plotting Small Samples with R.

R has a comprehensive set of routines for visualization of data. Here we show some of its capabilities on  a small sample Y containing positive values.


library("vioplot")

X <- c(rnorm(50, 10, 3),rnorm(50, 15, 5))
Y <- c(runif(10, 5, 20))


par(mfrow=c(3,3))

pie(Y, main="Small sample", sub="A Pie Chart")
dotchart(Y, main="Small sample", sub="A Dot Chart")
stripchart(Y, main="Small Sample", sub="A Strip chat")

barplot(Y, main="Small sample", sub="A vertical Bar Chart")
barplot(Y, horiz=T, main="Small sample", sub="A Horizontal Bar Chart")

boxplot(Y, main="Small sample", sub="A box plot")
vioplot(Y)  # package vioplot
title <- "A vio plot"

plot(Y, type="p", main="Small Sample", sub="A Scatterplot")
plot(Y, type="l", main="Small Sample", sub="A Line plot")


First a few observations:

  1.The vioplot function does not allow seting the main title or sub title. Instead you should call title("title of graph") after the call to vioplot.

  2. Pie charts are not well loved by statisticians, they cannot represent negative values and its difficult to judge magnitudes.



Here is a screenshot of the small sample plots.

Next time we will present plots which can handle a larger number of data points.